416 GREAT REFRACTING TELESCOPE AT ELCHIES, IN MORAYSHIRE. 



nitude, No. 18, b Aquilse, discovers a change of magnitude in its B ; No. 7, 222 

 Arietis, if not the same in its A, at least rectifies a standard book on a test star ; 

 and the new small stars found around Nos. 3, 17, 20, 21, 24, 25, and 27, promise 

 to form useful points of reference, touching proper-motion determinations by 

 equatorial observers, such as are much required towards the full understanding, 

 or for the due classification of those stars, according to character. 



In Colour something more has been done ; for variations of colour in several 

 stars have been shown to be real and extensive, sometimes affecting both mem- 

 bers, and sometimes one only, of a double-star; and in No. 13, the changes are 

 regularly periodical. This case is new, and, if true also, must be of extraordinary 

 importance even in days of " spectrum analysis." It is new, for putting on one 

 side the changing colours of those strange isolations, the temporary stars, and 

 certain degrees of change in tint of stars variable in brightness at the time that they 

 are varying,— the best astronomical writers, from Sir John Herschel downwards, 

 seem to allude to the colours of stars as so many " statical " facts ; things to be 

 once well determined by eye or by spectrum, and then stored away in classified 

 lists, or printed in books for permanent reference like the colours of minerals or 

 precious stones. But with our No. 13, or 95 Herculis, its two stars are neither 

 temporary nor variable in magnitude, and yet they change and change exceed- 

 ingly in colour ; their contrasted colours are not the effect of the light of one 

 powerful star dominating that of the other to the eye as a complementary tint, 

 — because the two stars are almost exactly of equal brightness ; and again the 

 bluer colour of one is not an effect of greater distance, because the stars are 

 proved, and now proved for the first time, to be binary. Hence, take them all 

 in all, the two stars of 95 Herculis, do seem to be really periodical changers of 

 cosmical colour from causes inherent in themselves, or connected with their own 

 region of space ; appearing thereby to indicate the propriet}'-, both of a date being 

 attached to ever}^ future observation of star-colour when made, and, on the 

 strength of the analogy between all true stellar orbs, a record of spectrum 

 analyses of ihe light of our own sun being kept up regularly, as a check upon 

 periodical or secular variations in the quality or material of his light.* 



* Some further inquiries may also be subserved by the Elchies colour observations ; which, if 

 unusually few in number, have been more than usually attended to touching the character of the 

 double stars concerned, or the difierence of the distance of their members from our system. 



There is undoubtedly an ether filling space, say most scientific men ; well then, if so, what is its 

 colour by transmitted light ? Star observations are peculiarly adapted to this end, and the colours 

 which we recognise in all stars may partly belong to this medium, whose colour, too, and composition 

 may be varied in particular regions. First, then, let us ascertain if there is any constant feature of 

 colour dependent on distance. Now the two nearest well-determined stellar systems are those of a 

 Centauri and 61 Cygni ; and Sir J. Herschel has remarked their strong yellow colour, stronger in 

 the small, than the large, component in either case ; if, then, those stars are really white, but appear 

 yellow to us, they give in so far a quadruple proof of the medium which extends between them and 

 ourselves, being yellow by transmitted light ; by no means an extraordinary result, if, according to 

 recent mathematicians, the atmosphere of the earth thins away and extends indefinitely into the 

 planetary and stellar spaces. 



I 



